Metabolomic analysis of urine and serum in Parkinson’s disease

Objective To investigate the metabolic profile of serum and urine samples from 23 female patients with Parkinson’s disease (PD) and 23 age and sex-matched controls. Methods We used gas chromatography coupled to mass spectrometry to detect metabolites (approximately 1,600 in total), then supervised statistical analysis (using projection to latent structures discriminant analysis) to study the differences between control and PD samples. Results Supervised statistical analysis yielded models that possessed statistically significant predictive value for blind samples on the basis of the metabolic profile of urine but not of serum. However, whilst no individual biomarkers were identified, suggesting that any metabolic disturbance associated with PD is comparatively minor, a multivariate metabolic signature associated with PD was identified in urine. Interpretation There is a relatively subtle, yet distinct, metabolic signature of PD present in the urine of patients with early disease. The signature may itself act as a useful biomarker for PD, although larger studies will be required to validate our present findings.